SYNOPSIS Energy/power input in a flotation cell is an important parameter which, if optimised, can increase the flotation rate. The optimum energy/power input within a flotation cell is still a matter of conjecture and there is a need for a better understanding of the effect of energy input on flotation kinetics. This study investigates the effect of energy/power input on flotation kinetics in an oscillating grid flotation cell (OGC). The OGC decouples the processes of solid suspension and bubble generation as well as producing relatively isotropic and homogeneous turbulence with zero mean flow. Due to this, oscillating grids provide a potentially ideal environment for investigating the effects of energy input on flotation kinetics, which cannot be achieved in a mechanical flotation cell. The first objective of this thesis was to determine the effect of energy/power input on the flotation kinetics of sulphide minerals (galena, pyrite & pentlandite) and oxide minerals (apatite & hematite) in a laboratory scale oscillating grid flotation cell. The second objective was to compare the results from the laboratory OGC to comparative studies in the flotation literature and to fundamental models for particle-bubble contacting. The third objective was to determine whether the experimental results from the laboratory OGC are consistent with those from a pilot-scale OGC operating on a platinum ore. Galena, pyrite, pentlandite (-150 μm), apatite (-650 μm) and hematite (-75 μm) were floated in the laboratory OGC at energy inputs from 0.1 to 5.0 W/kg, using 0.13, 0.24, 0.58 and 0.82 mm bubble sizes (d₁₀), and at three collector dosages. Platinum ore (-75 μm) was floated in the pilot-scale OGC at energy inputs from 0 to 2.5 W/kg, using 0.71 and 1.47 mm bubble sizes (d₁₀). The effect of energy input on flotation kinetics was interpreted through trends in experimental flotation rate constants, simulated flotation rate constants and attachment-detachment flotation rate constants. Here, simulated flotation rate constants were calculated using a literature fundamental model for flotation in turbulent systems. This model is based on suitable expressions for the collision frequency, collision efficiency, attachment efficiency and stability efficiency, Attachment-detachment flotation rate constants were calculated using a kinetic model which allows for the two separate processes of bubble-particle collision/attachment and detachment. This model is based on kinetic expressions using empirical correlations for the attachment and detachment rate constants. Experimental flotation results show that the effect of energy input on the flotation rate is strongly dependent on the particle size and particle density and less dependent on bubble size and contact angle. Flotation rates generally increase with increasing particle size, decreasing bubble size and increasing contact angle, as is commonly found in the literature. Increasing energy input generally leads to an increase in the flotation rate for fine particles, an optimum flotation rate for intermediate particles and a decrease in the flotation rate for coarse particles. The optimum in the flotation rate for minerals with higher density is at a lower energy input than that for lower density minerals. The changes (increases/decreases) in the flotation rate with increasing energy input are very large for most of the conditions, indicating that this is an important parameter in flotation. Pilot scale results generally support the trends observed in the laboratory OGC. These findings are attributed to the effect of energy/power input on bubble-particle collection which is a balance between two competing effects, those of bubble-particle collision/attachment and those of bubble-particle detachment. Increasing energy input generally leads to significant increases in the flotation rate of fine particles, due to increased bubble-particle collision/attachment. Increasing energy input generally leads to an optimum flotation rate for intermediate particles, due to a combination of increased bubble-particle collision/attachment and detachment. For coarse particles, increasing energy input leads to significant increases in bubble-particle detachment. The relationship between the flotation rate and energy input is often described as k ɛᴺ, in the absence of significant bubble-particle detachment. The typical values of N are in the range of 0.44-0.75 for theoretical studies and 0.7-1 for experimental studies. The values of N found in the current study are in the range of 0.7-1, which suggests that bubbleparticle collision/attachment has a stronger dependence on energy input than theory suggests. Simulated flotation results for fine particles compare well to the experimental data in terms of both trends and magnitude. This suggest that the turbulent collision model used is appropriate for fine particles. For intermediate particles there are differences between the simulated flotation rate constants and the experimental data, primarily in terms of trends. For coarse particles there are very large differences between simulated flotation rate constants and the experimental data. This is attributed to under prediction of the collision frequency/efficiency and incorrect prediction of the stability efficiency. Here, the stability efficiency is considered to be under predicted at low energy inputs and over predicted at high energy inputs. This suggests that the stability efficiency has a much stronger dependence on energy input than theory suggests. Attachment-detachment results show that the attachment rate constant has a stronger dependence on energy input than theory suggest, supporting finding from the experimental results and simulated results for coarser particles. In addition, the detachment rate constant has a much stronger dependence on energy input than theory suggests, supporting findings from both the experimental and simulated results. Based on the objectives of this study and literature reviewed, the following hypotheses were made at the outset 1) Increasing energy/power input will increase the rate of flotation of fine particles but will result in an optimum for intermediate and coarse particles. The position of this optimum will depend on the particle density, bubble size and contact angle. 2) Fundamental models based on the RMS turbulent velocity will be appropriate for describing flotation kinetics as turbulence in the oscillating grid cell is relatively homogeneous and isotropic and 3) Trends in flotation results for a laboratory and pilot-scale oscillating grid flotation cell will be comparable as the distribution of turbulence in OGCs at equivalent specific power inputs is scale independent. Hypothesis 1 was found to be valid for both fine and intermediate particles, but for coarse particles increasing energy input resulted in sharp decreases in the flotation rate. In addition, the increase in the flotation rate with increasing energy input was found to be more dependent on the particle size and particle density than the bubble size and contact angle. Hypothesis 2 was found to be valid for fine particles but not for intermediate or coarse particles. Here, it was found that the processes of bubble-particle collision/attachment and detachment have a stronger dependence on energy input than theory suggests. Hypothesis 3 was supported by general trends in results for the laboratory and pilot-scale oscillating grid flotation cells, but was not convincingly demonstrated.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/28162 |
| Date | January 2018 |
| Creators | Safari, Mehdi |
| Contributors | Deglon, David A |
| Publisher | University of Cape Town, Faculty of Engineering and the Built Environment, Centre for Minerals Research |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Doctoral Thesis, Doctoral, PhD |
| Format | application/pdf |
Page generated in 0.0026 seconds